Assisted Intelligent Routing for Minimalistic Connected Object Networks

1. A method, comprising: collecting, by a distributed intelligence agent (DIA), local state information from a plurality of minimalistic connected objects (MCOs) in a computer network, the local state information for each MCO comprising a corresponding neighbor list and a selected next-hop for the respective MCO, wherein one or more of the MCOs are configured to select their next-hop without any self-optimization; analyzing, by the DIA, a current routing topology, which is the combined result of the selected next-hops, in comparison to a computed optimal routing topology to determine whether to optimize the current routing topology; and in response to determining that the current routing topology should be optimized, transmitting a unicast routing instruction from the DIA to one or more individual MCOs to instruct those individual MCOs how to optimize the current routing topology.

2. The method as in claim 1 , further comprising: receiving updated local state information from the MCOs at the DIA; analyzing a corresponding updated current routing topology; and transmitting additional unicast routing instructions in response to determining that the updated current routing topology should be further optimized.

3. The method as in claim 1 , further comprising: computing the optimal routing topology based on one or more constraints.

4. The method as in claim 1 , further comprising: computing the optimal routing topology based on meeting one or more required service level agreements (SLAs).

5. The method as in claim 1 , wherein determining whether to optimize the current routing topology is based on at least one factor selected from a group consisting of: a total cost differential between the current routing topology and the optimized routing topology; a maximum path cost difference between the current routing topology and the optimized routing topology; a path within the current routing topology not meeting one or more required service level agreements (SLAs); and a number of changes required to optimize the current routing topology.

6. The method as in claim 1 , wherein the transmitted routing instruction comprises a new next-hop selection for the respective individual MCO.

7. The method as in claim 1 , wherein the local state information comprises at least one of local link cost, local link reliability, local link delay, local link expected transmission count (ETX), and a link-based received signal strength indication (RSSI).

8. An apparatus, comprising: one or more network interfaces to communicate in a computer network; a processor coupled to the network interfaces and adapted to execute one or more processes; and a memory configured to store a process executable by the processor, the process when executed operable to: collect local state information from a plurality of minimalistic connected objects (MCOs) in a computer network, the local state information for each MCO comprising a corresponding neighbor list and a selected next-hop for the respective MCO, wherein one or more of the MCOs are configured to select their next-hop without any self-optimization; analyze a current routing topology, which is the combined result of the selected next-hops, in comparison to a computed optimal routing topology to determine whether to optimize the current routing topology; and in response to determining that the current routing topology should be optimized, transmit a unicast routing instruction to one or more individual MCOs to instruct those individual MCOs how to optimize the current routing topology.

9. The apparatus as in claim 8 , wherein the process when executed is further operable to: receive updated local state information from the MCOs; analyze a corresponding updated current routing topology; and transmit additional unicast routing instructions in response to determining that the updated current routing topology should be further optimized.

10. A method, comprising: participating, by a minimalistic connected object (MCO) with a plurality of MCOs in a computer network, in a build of a current routing topology without any self-optimization, an initial instance of the current routing topology having loop prevention, but having no objective function, no mechanisms for local repair, and no propagating advertisement messages; transmitting, from the MCO to a distributed intelligence agent (DIA), local state information comprising a corresponding neighbor list and a selected next-hop for the MCO for the current routing topology; receiving a unicast routing instruction from the DIA at the MCO that instructs the MCO how to optimize the current routing topology; and re-selecting a next-hop for an updated current routing topology according to the routing instruction from the DIA.

11. The method as in claim 10 , further comprising: transmitting updated local state information from the MCO to the DIA in response to the updated local state information being a significant change from the previously transmitted local state information.

12. The method as in claim 10 , wherein the local state information comprises at least one of local link cost, local link reliability, local link delay, local link expected transmission count (ETX), and a link-based received signal strength indication (RSSI).

13. An apparatus, comprising: one or more network interfaces to communicate in a computer network; a processor coupled to the network interfaces and adapted to execute one or more processes; and a memory configured to store a process executable by the processor, the process when executed operable to: participate, with a plurality of minimalistic connected objects (MCOs) in a computer network, in a build of a current routing topology without any self-optimization, an initial instance of the current routing topology having loop prevention, but having no objective function, no mechanisms for local repair, and no propagating advertisement messages; transmit, to a distributed intelligence agent (DIA), local state information comprising a corresponding neighbor list and a selected next-hop for the apparatus for the current routing topology; receive a unicast routing instruction from the DIA that instructs the apparatus how to optimize the current routing topology; and re-select a next-hop for an updated current routing topology according to the routing instruction from the DIA.

14. A system, comprising: a plurality of minimalistic connected objects (MCOs) in a computer network, the MCOs configured to select their respective next-hop in a current routing topology without any self-optimization, the MCOs further configured to transmit, to a distributed intelligence agent (DIA), local state information comprising a corresponding neighbor list and the respective selected next-hop; and a DIA configured to analyze the current routing topology, which is the combined result of the selected next-hops, in comparison to a computed optimal routing topology to determine whether to optimize the current routing topology, and also to transmit, in response to determining that the current routing topology should be optimized, a unicast routing instruction to one or more individual MCOs to instruct those individual MCOs how to optimize the current routing topology; wherein those individual MCOs are further configured to receive the unicast routing instruction from the DIA, and to re-select a next-hop for an updated current routing topology according to the routing instruction from the DIA.

15. The system as in claim 14 , wherein the DIA is further configured to: receive updated local state information from the MCOs at the DIA; analyze a corresponding updated current routing topology; and transmit additional unicast routing instructions in response to determining that the updated current routing topology should be further optimized.

16. The system as in claim 14 , wherein the DIA is further configured to compute the optimal routing topology based on at least one of the following: one or more constraints; and meeting one or more required service level agreements (SLAs).

17. The system as in claim 14 , wherein determining whether to optimize the current routing topology is based on at least one factor selected from a group consisting of: a total cost differential between the current routing topology and the optimized routing topology; a maximum path cost difference between the current routing topology and the optimized routing topology; a path within the current routing topology not meeting one or more required service level agreements (SLAs); and a number of changes required to optimize the current routing topology.

18. The system as in claim 14 , wherein the transmitted routing instruction comprises a new next-hop selection for the respective individual MCO.

19. The system as in claim 14 , wherein the local state information comprises at least one of local link cost, local link reliability, local link delay, local link expected transmission count (ETX), and a link-based received signal strength indication (RSSI).

20. The system as in claim 14 , wherein the MCOs are further configured to: participate in a build of a current routing topology without any self-optimization, an initial instance of the current routing topology having loop prevention, but having no objective function, no mechanisms for local repair, and no propagating advertisement messages.

21. The system as in claim 14 , wherein the MCOs are further configured to: transmit updated local state information to the DIA in response to the updated local state information being a significant change from the previously transmitted local state information.